The post-reload probe escalated on any status the app does not accept, and a
forced OpenResty task redeploy costs a measured 40.32s. In four CI job logs it
fired 11 times per job on 502 and 500 only, never on 404 and never on a
connection failure.
Those statuses refute the escalation rather than justify it. OpenResty emits
5xx after it has matched a server block and proxied, so they prove the vhost is
loaded and the reload took effect. The default server cannot emit them at all:
it returns 404 on port 80 and rejects the handshake on 443, and it carries no
proxy_pass. In eight of eight round-one escalations the upstream app had not
been deployed yet when its vhost was probed, the deploy trigger following the
502 by up to 52 seconds.
The log also holds a control. A full 40s redeploy completes three to six
seconds before every one of the 28 post-reload probes, notified by another
pending restart in the same flush. Seventeen then pass and eleven do not, same
proxy, same recency, so the differentiator is the backend. The escalation
answers that by running the same redeploy again, and nothing re-probes after it.
Narrow the trigger to the two signatures that do prove a stale configuration,
plus the case where the reload never landed because the container was
mid-recreate, which reload.yml tolerates. Worth 438 to 470 seconds per job.
The status test also gains the connection-failure case. Today the
`status is defined` clause excludes it, so the one symptom a genuinely stale
config produces could never fire the escalation at all.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* scripts/lint/autoformat.sh:
- rename run_sort_claude_settings worker to run_claude (matches the
renamed scripts/lint/claude.sh).
- drop the EXIT trap that wiped status_dir; cleanup happens explicitly
after `wait` returns, so a parallel worker that interrupts the
parent never races a vanished status dir.
- each worker now captures its tool's exit code and writes one of
OK / SKIP / FAIL. Tolerated `expected violations remain`-style exits
(markdownlint-cli2 rc≤1, ansible-lint rc≤2) still count as OK; only
real failures count as FAIL.
- all workers run to completion regardless of individual tool failures
(so the user always sees the full picture); after the summary, the
script exits 1 if any worker reported FAIL.
* scripts/lint/sort_claude_settings.sh -> scripts/lint/claude.sh
(history preserved via git mv); update test_claude_settings_sorted.py
fix-hint accordingly.
Plus working-tree drift across roles/, group_vars/, inventories/,
requirements/, and .pre-commit-config.yaml accumulated from the
make-test setup phase regenerating generated artefacts during recent
test runs.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Refactor proxy/webserver configuration variables to a consistent webserver_* naming scheme across roles. Replace legacy variables like proxy_extra_configuration, client_max_body_size, vhost_flavour, location_ws and ws_port with webserver_extra_configuration, webserver_client_max_body_size, webserver_vhost_flavour, webserver_websocket_location and webserver_websocket_port. Update NGINX vhost and location templates (html, upload, ws, basic, ws_generic) as well as callers (sys-front-inj-all, sys-stk-front-proxy, various web-app-* and web-svc-* roles) to use the new naming.
Tighten docker-compose Git repository handling by making docker_git_repository_pull depend on docker_git_repository_address being defined, a string and non-empty. This avoids accidental Git operations when the repository address is unset or of the wrong type.
Refactor the BigBlueButton role structure and fix deployment bugs: introduce 01_core.yml to orchestrate docker/proxy setup, database seeding, websocket map deployment, docker-compose overrides and admin/bootstrap logic in a single once-executed entrypoint. Rename supporting task files (02_docker-compose.yml, 03_administrator.yml, 04_dependencies.yml) and update tasks/main.yml to delegate via include_tasks with run_once_web_app_bigbluebutton. Improve Greenlight admin creation behavior by treating the 'Email has already been taken' error as a non-fatal, unchanged outcome and running user:set_admin_role as a fallback, both for the primary password and the OIDC starred-password path.
Also standardize vhost flavour selection for services like Mailu, Discourse, CDN, Collabora, Coturn, OnlyOffice, Simpleicons and web-svc-logout by explicitly passing webserver_vhost_flavour where needed and aligning client_max_body_size and websocket configuration with the new webserver_* variables.
Reference: ChatGPT conversation https://chatgpt.com/share/6931c530-bba8-800f-9997-dd61dc1d497b
Integration tests added/updated:
- tests/integration/test_filters_usage.py: AST-based detection of filter definitions (FilterModule.filters), robust Jinja detection ({{ ... }}, {% ... %}, {% filter ... %}), plus Python call tracking; fails if a filter is used only under tests/.
- tests/integration/test_filters_are_defined.py: inverse check — every filter used in .yml/.yaml/.j2/.jinja2/.tmpl must be defined locally. Scans only inside Jinja blocks and ignores pipes inside strings (e.g., lookup('pipe', "... | grep ... | awk ...")) to avoid false positives like trusted_hosts, woff/woff2, etc.
Bug fixes & robustness:
- Build regexes without %-string formatting to avoid ValueError from literal '%' in Jinja tags.
- Strip quoted strings in usage analysis so sed/grep/awk pipes are not miscounted as filters.
- Prevent self-matches in the defining file.
Cleanup / removal of dead code:
- Removed unused filter plugins and related unit tests:
* filter_plugins/alias_domains_map.py
* filter_plugins/get_application_id.py
* filter_plugins/load_configuration.py
* filter_plugins/safe.py
* filter_plugins/safe_join.py
* roles/svc-db-openldap/filter_plugins/build_ldap_nested_group_entries.py
* roles/sys-ctl-bkp-docker-2-loc/filter_plugins/dict_to_cli_args.py
* corresponding tests under tests/unit/*
- roles/svc-db-postgres/filter_plugins/split_postgres_connections.py: dropped no-longer-needed list_postgres_roles API; adjusted tests.
Misc:
- sys-stk-front-proxy/defaults/main.yml: clarified valid vhost_flavour values (comma-separated).
Ref: https://chatgpt.com/share/68b56bac-c4f8-800f-aeef-6708dbb44199
- Renamed role `srv-tls-core` → `sys-svc-certs`
- Renamed role `srv-https-stack` → `sys-stk-front-pure`
- Renamed role `sys-stk-front` → `sys-stk-front-proxy`
- Updated all includes, READMEs, meta, and dependent roles accordingly
This improves clarity and consistency of naming conventions for certificate management and proxy orchestration.
See: https://chatgpt.com/share/68b19f2c-22b0-800f-ba9b-3f2c8fd427b0
kevinveenbirkenbachandClaude Opus 5